Page 136 - AJWEP-22-4

Wang, et al.

Table 1. Descriptive statistics digital transformation, while average humidity has an
Variable n Mean SD Min Max inhibitory effect.
name
Digital 1 12,908 47.16 87.83 0 1264 4.3. Robustness test
To ensure the results are robust, firstly, the measure of
Digital 2 12,908 1.051 1.24 0 6.139 digital transformation was replaced, drawing on Yuan
Extreme 12,908 35.52 39.64 0 173 et al. and Zhao et al. to increase the frequency of
39
40
Size 12,908 21.98 1.288 19.14 26.45 words related to digital transformation of enterprises to
Lev 12,908 0.426 0.203 0.027 0.908 99, to re-measure the degree of digital transformation
ROA 12,908 0.038 0.065 -0.373 0.257 of enterprises; the results are as shown in column
Board 12,908 2.141 0.210 1.099 2.833 (1) of Table 3. The extreme temperature variable is
ListAge 12,908 1.947 0.925 0 3.401 significantly positive. Second, the extreme temperature
in the benchmark regression was calculated using
Indep 12,908 35.70 8.567 0 60 an absolute threshold, which is simple, but also has
Dual 12,908 0.250 0.433 0 1 defects. To avoid the bias generated using the fixed
41
TOP1 12,908 35.06 15.12 8.020 75.84 threshold method, this study re-measured the urban
BM 12,908 0.643 0.241 0.064 1.246 extreme temperature index using the relative threshold
SOE 12,908 0.401 0.490 0 1 and expanding the fixed threshold method, respectively,
Big4 12,908 0.060 0.237 0 1 and carried out the test. Column (2) of Table 3 expands
Opinion 12,908 0.966 0.180 0 1 the range of extreme temperatures to above 30°C and
below 0°C, and the results are still significant. Column
IC 12,908 648.9 127.9 0 999.8 (3) of Table 3 draws on Alexander et al.’s percentile
GDP 12,908 91662 56665 2093 467749 interpolation and refers to Pan and Zhai to calculate
42
43
EDU 12,908 33.23 29.02 1 93 the cumulative number of days of extreme temperatures
ENV 12,908 0.262 0.089 0 0.879 using the 95% percentile of daily high temperatures
Wind 12,908 5.359 1.043 2.222 8.967 and the 5% percentile of daily low temperatures,
Wet 12,908 70.93 9.098 35.63 84.46 respectively, and the results showed that the extreme
Sun 12,908 1935 428.5 752.4 3386 temperatures still have a facilitating effect on the digital
transformation of enterprises. Up to this point, possible
bias due to variable measurement is ruled out.
4.2. Baseline regression results Considering the early development of digital
economy in Zhejiang Province, the national leader,
Table 2 reports the results of multiple regressions of as well as the persistence of the role of temperature,
extreme temperatures on firms’ digital transformation, we deleted the samples of listed companies based in
where industry-fixed effects and year-fixed effects are Zhejiang Province, as well as lagged one period of the
not included in columns (1) and (3). As seen from enterprise digital transformation variables; the results
the regression results, the coefficients of the extreme are shown in the columns (4) and (5) of Table 3. The
temperature variables are significantly positive under extreme temperature variables are significantly positive
all models, indicating that the frequency of extreme at the 1% level, which further excludes the sample bias
temperatures promotes the digital transformation of that may be generated due to the differences in time and
manufacturing firms, and Hypothesis 1 is basically geography.
verified. The lagged terms of extreme temperature are
all significantly positive, indicating the continuity of 5. Mechanistic analysis
the effect of extreme temperature on enterprise digital
transformation. Company size, shareholder size, and In light of the reliable results from the baseline
the number of years on the market have a facilitating regression, we further tested the mechanism. According
effect on enterprise digital transformation, which is to the previous analysis, extreme temperatures promote
basically consistent with existing research. Turning to the digital transformation of enterprises by increasing
the climate variables, the coefficients of the light hours enterprise costs and reducing enterprise efficiency. To
and average wind speed variables are significantly test hypothesis 2, we measured enterprise cost pressure
positive and have a facilitating effect on enterprise from the perspectives of wage structure, cost growth

Volume 22 Issue 4 (2025) 128 doi: 10.36922/AJWEP025210166